We propose to further develop and maintain an open source software toolbox, 'EEGLAB' for analysis and visualization of human brain dynamics from electroencephalographic (EEG) data. EEGLAB offers the ever growing cognitive neuroscience research community an extensible, open source software platform for applying modern signal processing and dynamic visualization methods in cognitive neuroscience research. EEGLAB is designed to facilitate a major evolutionary step now taking place in the field of human, electrophysiology. Long dominated by simple time-locked event-related potential (ERP) averaging methods, human EEG research is increasingly making use of modem signal processing advances including time/frequency analysis, source localization onto magnetic resonance images of the head, three-dimensional and animated visualization methods, and independent component analysis (ICA), methods that are applied not only to trial averages but directly to the high-dimensional single-trial data. Research in our laboratory and elsewhere has shown that much unexploited information about human brain dynamics contained in these data is accessible to these methods. Yet, despite the ever-increasing affordability of high-speed computational resources that should allow every EEG laboratory to perform sophisticated analyses on their data, the use of modern signal processing tools by EEG researchers is still rare. We thus propose to further develop and maintain an established public domain, open source software toolbox, EEGLAB 'http://sccn.ucsd.edu/eeglab/) running under the cross-platform MATLAB programming environment (The Mathworks, Inc.). EEGLAB encompasses functions at three levels of user sophistication: a purely graphic interface for new or casual users, high- and low-level scripting methods for experienced users, plus a software plug-in facility that allows innovators to easily introduce new analysis approaches to the EEG research community.